Multinode Spectrum Sensing Based on Energy Detection for Dynamic Spectrum Access

Abstract

Sharing of the frequency spectrum between licensed primary users and unlicensed secondary users (SUs) requires reliable detection of spectrum occupancy by the SUs. Due to fading, single terminal detection is unreliable and results in a high probability of missed detection. This problem is solved by applying cooperative detection. In this paper two novel energy-based cooperative detection methods using weighted combining for Dynamic Spectrum Access are presented and analyzed. Weighting is based on the local mean SNR and the optimum log-likelihood ratio. Simulation results show a substantial improvement for the proposed weighting methods compared to equal gain combining and hard decision combining.